I posted this in general because it's a very frequently discussed subject that's continually revisited. However, if it's felt that it belongs elsewhere, please move it.

Let’s take a look at Residential Security Containers, hereafter referred to as RSC’s, and safes. First, let me state that there is no legal provision that I know of that regulates what may, or may not, be called a safe. In other words, I can take Scotch Tape and business cards, tape them into a little box, put a pin across the corner, and legally sell it as a “safe”. Shame on you if you pay good money for it, but it’s an excellent example of caveat emptor.

The best, in my opinion, rater of security containers is Underwriter’s Laboratories. However, simply because a container has a U.L. label it does not mean you are looking at a true safe. Please be aware that there are several rating categories in the U.L. system. I’ll provide my synopsis of the pertinent areas, but it’s beyond the scope of this thread to quote the entire manual. Feel free to go to the official U.L. website & wade through the verbiage if you care to. However, also be aware that U.L. is the only tester that I know of that openly publishes their test parameters. And that’s a significant point when attempting to do an honest comparison of manufacturer’s claims.

The security categories I’ll be talking about are the RSC, construction, attack resistance, and thermal protection. Simply put, an RSC designation means that sheet metal has been used in the construction of the security container. Beyond that, the RSC label means that the container has been tested to resist physical attack by one person, using hand tools (commonly meant to mean a hammer and large screwdriver) with no lever length exceeding 18 inches, for 5 minutes. Many import units now will take the body sheet metal, bend it into the interior, fold it three or four times, & use that as the frame which the bolts lock up behind. What happens when you fold sheet metal? Rhetorical question, folding sheet metal weakens it. It’s possible to view an excellent example of this type of construction by accessing the YouTube video “Security On Sale”. Therein two young men flop an RSC on it’s back & pop the door in one minute and 42 seconds I think. A careful examination of the “crook’s” movement patterns will reveal that they obviously weren’t rehearsed. A frequently heard objection to the methodology of the YouTube video is that if the RSC were bolted upright, the “crooks” couldn’t have done what they did. Please. It’s true that if the RSC were upright, it would have taken them longer, but the point of the video is valid. A major reason I’d think that the RSC was flopped is that most people will sit through a couple of minutes of sales video, but literally won’t sit still for a ten minute video pitch. Nonetheless, no RSC, regardless of other considerations, is high-grade protection from physical attack.

U.L. will not rate a security container as a safe unless all six exterior surfaces are at least ” inch thick plate steel. Which brings us to construction ratings. A “B” rate safe will have, at a minimum, quarter inch plate used throughout its exterior construction. The better one’s will have a door of ” thick steel plate. And the steel will not be mild rolled plate. Better “B” rated safes will use something like A36 high tensile tool steel for their construction. There are further construction ratings such as C, E, F, and beyond. A “C” rated unit will have 5 exterior surfaces of ” plate and a 1” thick plate door. An “E” rated safe will have 1” thick exterior surfaces with the door being 1.5” thick. The “F” rated safe will incorporate a layer of manganese steel laminated in the exterior surfaces to deter torch attack. Note that the construction ratings of “B” etcetera are not directly comparable to attack resistance ratings of TL, TR, etc. However, careful examination of the two rating systems does allow justifiable conclusions to be drawn.

An attack rating label, such as TL15, certifies that the door will withstand an attempted forced opening by tools for 15 continuous minutes. That’s the door only. If a unit is rated as TL15 X 6, then the entire body of the safe (all six sides) is certified to withstand the attack for 15 minutes. The common ratings concerning tool attack are TL15 and TL30, for 15 and 30 minutes of successful resistance. A TR rating is a resistance to a torch attack. The same time and side parameters apply to the nomenclature. So, if you’re in possession of the TLTR30 X 6 safe, you’ve got one helluva unit. However, that type of unit usually carries a price tag so as to preclude ownership by private individuals. In other words, they are very expensive.

When it comes to thermal protection ratings for home use, the only one to pay any attention to is the U.L. one-hour certification. You will almost certainly run across several different “systems” if you go shopping for a “safe”. There’s the Omega Laboratories label, the Pyro 3000, the BTU certification, and others. Until you know the test parameters of each “system”, it’s impossible to draw any accurate comparisons. What’s more, the only tester that I know of that does publish the aforementioned parameters is Underwriters Laboratories. A thumbnail description of the U.L. process is this: The test safe goes into the furnace, the door is shut & the fire is lit. The one-hour test timer only starts when the furnace temperature rises to 1700 degrees f. At the end of the hour, anything that’s going to pass the test will have its internal test transponder reporting an interior temperature of something like 270 – 280 f. That’s not the end of the test however. The burn in the furnace is stopped, but the furnace remains sealed until it’s internal temperature falls to laboratory ambient, 68 f. At no time can the internal temperature of the test container rise above 350f for the entire test. The 350f figure is the temperature determined by U.L. that data is recoverable from good quality paper. As an aside, the paper quality parameters cover some few pages themselves. Not, “can’t put it down” reading.

When confronting other thermal rating figures, you should ask yourself some questions: “What’s the rise time to test temperature?” “Did the test period timer start when the gas was lit, or when test temperature was reached?” “Was the container placed in the test furnace in the upright position that it would be in my home?” “Did the test include the furnace cool-down time?” Until you know the answers to questions such as these, it’s just not possible to say that the container provides X number of minutes protection at Y temperature, and be able to draw a meaningful conclusion from the numbers.


OK, thermal protection is not rocket science. RSC’s use sheet metal for body construction & they will usually use either sheet rock, i.e. gypsum wallboard, or a foam filler for the insulating medium. Sometimes gypsum board sans paper will be called “fire rock”, don’t be fooled, its the same stuff. In any case, my point is that if you put thicker and denser material between the heat source & what you want saved, you get better protection with plate steel and concrete than with sheet metal and gypsum. The thermal protection offered by virtually all RSC’s is minimal in my experience.

When it comes to container locks, the only two real choices these days are digital, or manual combination. But, there are sub-categories that make substantial differences in both the reliability and protection you get when you buy. Many manual combination dial locks on import RSC’s are what is known as a wheel-pack unit. They are relatively easy to make, offer the three number combination most of us are familiar with, and cannot be told from exterior examination from a high quality unit such as a U.S. made LaGard or S&G. However, if you want the combination changed, the LaGard and Sargent Greenleaf units are made to allow that relatively quickly and easily. The import wheel-pack units are frequently very much a PIA to change the combination. Speaking as a professional locksmith, I will charge you more to change the combination on a low-tech wheel-pack than I will to do the same job on a quality unit that’s made to have it done. There are inherent limitations due to the mechanics of combination dials that absolutely should be adhered to. They are: Each number should be at least 10 digits from the next, and the final number should not lie in the 80-0-20 quadrant. In example, a combination of 30-20-30 is OK, but a combo of 05-10-95 is sure to cause problems. But, there are thousands of possible combinations available to the user within those limitations. On the other hand, wheel-pack units may only offer you a few (12?) possible combinations, depending on the way the thing is built. In any case, I’ll strongly suggest that no end user who has not been trained to do so attempt to change their own manual combination dial. It will cost you major money if you try it & don’t get it exactly right.

Electronic locks offer some very real advantages to manual combination dials. There are also some very real drawbacks. An electronic keypad offers superior speed of opening, by several hundred percent, over the typical manual combo dial. Usually, the electronic keypad lock is far more user friendly when changing the combination also. Just in case though, I’ll offer this common sense rule: Do it with the door open, check it at least three times before trying it with the door closed. Failure to do so will result in your locksmith’s being able to afford a new toy. Another caveat: Always make sure that the electronic lock you’re considering has 1. an external battery, and 2. a burn memory. Locks without a burn memory rely on a capacitor to supply electricity to the memory card while you change the battery. Take more time than the capacitor has current & you may very well have a problem. A lock that has a burned in memory can have the battery removed while you travel around the world & still retain your particular combination upon your return & re-installing the battery.

Oh, and concerns about EMP’s, and super-crooks with monster lap top computers who will plug in to your lock & crack the combo in 28 seconds max are just not real world concerns. If you’re spending time worrying about those, then please also give due consideration to giant meteors and Yellowstone blowing up.

There you have it, my overview of the subject of safes for home use. Please PM me if you have questions about something that wasn’t covered in the article. Thanks for taking the time to read.

900F